基于PCA与SOM神经网络算法的客户价值细分模型研究

摘要

针对房地产客户价值管理问题,以某大型房地产企业的普通住宅业务为研究对象,构建了基于PCA与SOM神经网络算法的房地产客户价值细分模型.首先采用PCA主成份分析法将输入变量重组为线性不相关的综合指标,然后采用SOM神经网络算法对客户价值进行聚类分析,最后针对聚类结果,分析不同目标客户群的购买模式和价值特征,旨在为房地产企业的营销过程提供决策支持.%Facing real estate customer value management and taking common residence business of a large real estate enterprise as research object,this paper establishes a real estate customer value segmentation model based on PCA and SOM neural network.Firstly,input variables are reorganized into linear and irrelevant composite indicators in the PCA method.Then,a cluster analysis is made of customer value in the SOM neural network algorithm.Finally,with respect to cluster results,this paper analyzes purchasing patterns and value characteristics of different target customer groups so as to provide decision support for the marketing process of the real estate enterprise.